High-resolution seismic study of the Nahal Darga fan-delta ...The left-lateral Dead Sea transform (DST) is an active plate boundary between the Arabian and the African plates (e.g

EGU Stephan Mueller Special Publication Series, 2, 21–33, 2002c© European Geosciences Union 2002

High-resolution seismic study of the Nahal Darga fan-delta, DeadSea, Israel, with the aim to relate the surface and subsurfacetectonic structures

Y. Eyal1, I. Bruner 2, G. Cadan1, Y. Enzel3, and E. Landa2

1Department of Geological and Environmental Sciences, Ben-Gurion university of the Negev, Beer-Sheva, 84105, Israel2The Geophysical Institute of Israel, Lod, 71100, Israel3Institute of Earth Sciences, the Hebrew University of Jerusalem, Jerusalem, 91904, Israel

Received: 20 November 2000 – Revised: 21 November 2001 – Accepted: 28 November 2001

Abstract. Faulting and deformation of fan-delta deposits lo-cated near active tectonic zones can be attributed either tofailure due to basin-ward gravitational mass movement, or totectonic activity. Distinguishing between these two failuremechanisms is sometimes crucial when attempting to assessthe seismic risk and earthquake recurrence time of such ter-rains. In a case study described here, close spatial associ-ation between subsurface faults, determined in seismic pro-files, and overlying surface deformation features mapped inthe field, would imply tectonic activity as the cause for de-formation rather than basin-ward gravitational mass move-ment. The geologic and seismic study of the 400 m longNahal Darga fan-delta of Holocene age, located on the west-ern margins of the tectonically active Dead Sea transform, isused here as a test case for this debate. The studied fan-deltais comprised of two deformation zones. The Western defor-mation zone consists of NNE trending normal faults with upto 2 m offsets, mostly crossing part of the section and termi-nating below the surface. Only a few faults cross the entireHolocene section. The Eastern zone is characterized by manysmall normal faults, which are restricted to only a few bedsand NNE trending allochthonous bodies (slumps) that devel-oped due to tectonic activity. A shallow, high-resolution seis-mic study carried out in the Nahal Darga fan-delta, revealsthe existence of deep-seated faults extending up to about tenmeters below the surface. The location of the subsurfacefaults below the surface deformation zones suggests asso-ciation between the two and a tectonic origin for both. Itis concluded that detailed geological and structural mappingof fan-deltas coupled with a high resolution seismic surveyis an essential and reliable tool for the analysis of tectono-stratigraphy in fan-deltas.

Correspondence to:Y. Eyal ([email protected])

1 Introduction

The left-lateral Dead Sea transform (DST) is an active plateboundary between the Arabian and the African plates (e.g.Freund et al., 1970; Garfunkel et al., 1981). It extends fromthe northern end of the Red Sea divergent plate boundaryto the Bitlis convergent plate boundary in southern Turkey(Fig. 1). Along the DST many pull-aparts (e.g. the Dead Seaand Hula basins) and push-ups (e.g. Mount Hermon) devel-oped due to left and right stepovers of the transform trace(Garfunkel, 1981).

Nahal Darga drains the eastern part of the Judean Moun-tains and the Judean Desert located west of the northwest-ern part of the Dead Sea pull-apart (Fig. 1). The westernmargin of the Dead Sea is characterized by a north-south es-carpment, about 400–600 m high, comprised of a few topo-graphic steps due to step faulting. This escarpment is sub-parallel to the DST and is marked as ”border faults” of theDead Sea rift (Fig. 1). It resulted from up to 10 km downfaulting due to extension in the vicinity of the Dead Sea (e.g.Kashai and Croker, 1987) that occurred since the Pliocene tothe present. Prior to 5 Ma ago the DST was characterized bystrike-slip motion (Garfunkel, 1981). However, since thenthe trend of motion has been changed to an oblique one, asreflected by the change in Euler Pole, and also includes anextensional component along the DST resulting in rift devel-opment (Joffe and Garfunkel, 1987). Streams draining theeastern part of the Judean Mountains and Judean Desert intothe Dead Sea have incised narrow and very steep gorges, upto a few hundred meters deep, due to the high topographicescarpment. The slope of the riverbeds is steep and manywaterfalls developed up to ten kilometers west of the escarp-ment. The subsidence of the rift block and the depth of theDead Sea water in this area, about 400 m, created an idealsituation for the development of fan-deltas. The Nahal Darga

22 Y. Eyal et al.: High-resolution seismic study of the Nahal Darga fan-delta, Dead Sea, Israel

I II I

I II I

I I

I I I I I I I I I I I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

IIIII I I I I I I I I I I I I I I I I I

I II I

I II I

I II I

III

IIII

III I

I II I

IIII

I II I

II I

I II I

I II I

I II I

I II I

I II I

I II I

I I I I I I I I

II I

I

I I I I II I I I

35°30'E

31°45'E

32°00'

JERUSALEM

JERICHO

Jo

rda

n

Fa

ult

Jo

rda

nR

ive

r-450

-500

-550-600

-650

-700

DST

I I I I I I I I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII I I I I I

0 5kmN

IIIIIIIIEpicentral location of 1927 EQ(Shapira et al., 1993)with error ellipse

Active, mainly strike-slipfaults

Border faults, mainly normal,of the Dead Sea rift

Wadi

LEGEND

N . D a r g a

Jud

ean

Des

ert

Dea

d S

ea

S inaiSub-Plate R

ed

Se

a

A f r i c a n P l a t e

A r a b i a n P l a t e

30°E

Eurasia Plate

Med.Sea

30°N

A.

Fig. 1. Location map (after Niemi and Ben-Avraham, 1994) – a general tectonic setting of the Dead Sea transform (A – the study area).Insert – The general tectonic setting of the northern part of the Dead Sea (after Niemi and Ben-Avraham, 1994). The epicentral locationof the Jericho 1927 earthquake (Shapira et al., 1993) is about 2 km northeastward of the Nahal Darga fan-delta. The contour lines are thebathymetric contours of the Dead Sea.

Y. Eyal et al.: High-resolution seismic study of the Nahal Darga fan-delta, Dead Sea, Israel 23

fan-delta is the largest in the northwestern part of the DeadSea.

Fan-deltas are developed in areas where high relief is re-placed by low, such as a mountain range front, or marginsof deep-water bodies surrounded by high topographic areas.The heavy load carried by the high-energy water, character-izing steep intra-mountain rivers, is deposited as soon as thewater loses its potential energy when reaching the lowland.The load is differentially deposited according to the weightof the particles carried, the coarser near the river mouth andthe finer away from it.

Commonly, young fan-deltas such as that of Nahal Dargaconsist of poorly-consolidated or loose sediments. Displace-ment along exposed fault planes within such fan-deltas isvery common (see below) and may result from two mech-anisms:

1. landslides triggered by gravitational loading and move-ment toward the depositional basin, or

2. faulting due to tectonic activity.

It is important to determine which mechanism is responsiblefor given faults within young deposits, especially for areaswith evidence of historical earthquakes such as the Dead Sea,for the following reasons:

1. faults, with different amounts of displacement are com-mon in many fan-deltas which developed within or ad-jacent to tectonically active areas, e.g. along the DST(Zak and Freund, 1966; Bowman, 1995), in the Basinand Range Province (e.g. Bell and Katzer, 1987), andalong the San Andreas Fault (e.g. Sieh and Jahns, 1984);and

2. seismic hazard assessment is frequently based on recur-rence time of large earthquakes and this in turn dependson relating the faults and other deformation structuresto tectonic events.

The recurrence time of large earthquakes in the Dead Seaarea is commonly based on dating of rock units that weredisplaced along new or reactivated faults (e.g. Reches andHoexter, 1981; Marco and Agnon, 1995; Enzel et al. 2000).Thus, distinguishing between faults developed due to massmovement toward the basin and faults formed due to tec-tonic activity is crucial in an attempt to assess the seismicrisk and recurrence time. The existence of deep-seated sub-surface faults beneath a group of exposed faults in a fan-deltawould strongly support the development of the latter due totectonic activity, whereas the lack of such faults or the exis-tence of a shallow detachment surface may support a land-slide mechanism (e.g. Hoek and Bray, 1991).

In this paper we report the results of a geologic and seis-mic study in Nahal Darga fan-delta. It shows the associa-tion between surface faults and underlying subsurface faultsprecisely identified by detailed field mapping and a high-resolution seismic survey. This association suggests that theHolocene fan-delta surface faults are the extension of deep-seated larger faults.

2 Geological geomorphological setting

The Dead Sea transform, together with the North Anatolianand South Anatolian faults, is one of the main tectonicallyactive zones in the Middle East responsible for large andhazardous earthquakes (e.g. 1927 Jericho earthquake, and1999 Izmit earthquake along the DST and North AnatolianFault respectively). The pre-Holocene history of the tectonicactivity along the DST is based mainly on the analysis ofdisplaced rock units along faults associated with this activ-ity (e.g. Freund, 1965; Freund et al., 1970; Bartov, 1974;Garfunkel, 1981; Eyal et al., 1981), as well as other defor-mation structures, such as folds and pull-aparts or pushups.The latter resulted from the en-echelon and stepover patternsof the DST trace (Ben-Avraham, 1985; Freund, 1965; Gar-funkel, 1981; Heimann and Ron, 1987; Eyal et al., 1981).The Late Pleistocene to Recent history of seismic activity hasbeen based on a variety of features, e.g. faults within LatePleistocene and Holocene terraces (e.g. Bowman and Ger-son, 1986; Gardosh et al., 1990; Enzel et al., 2000); slumps inthe Jordan River sediments (Niemi and Ben-Avraham, 1994);analysis of deformed young sediments within trenches exca-vated across active faults (Reches and Hoexter, 1981; Amitet al., 1997); fluidized beds associated with surface fault-ing such as those found in the Lisan formation (Marco andAgnon, 1995); liquefaction structures (Enzel et al., 2000),destruction of archeological sites (e.g. Karcz et al., 1977;Ben-Menahem, 1991; Marco, et al., 1997), and monitoreddisplacement of key blocks due to seismic activity (Hatzor,1999).

Faulted scarps in Holocene rocks along the DST are abun-dant and were reported from various locations along the DST,for example: at the northwestern margin of the Gulf of Elat(Bowman and Gerson, 1986); within the municipal area ofthe city of Elat (Ginat et al., 1994); the southern Arava northof Elat (e.g. Zak and Freund, 1966; Gerson and Grossman,1991; Enzel et al., 1994); the Dead Sea area (e.g. Gardosh etal., 1990; Cadan, et al., 1996; Bowman, 1995); and the HulaBasin in northern Israel (Heimann et al., 1997).

Nahal Darga, as one of a number of streams draining theJudean Mountains and Judean Desert, crosses the fault es-carpment of the Dead Sea rift into the northwestern partof the Dead Sea basin. It terminates in an active fan-deltacomprising of poorly consolidated, Late Pleistocene andHolocene, fluvial, deltaic and lacustrine sediments (Enzel etal., 2000). The eastern part of the Nahal Darga fan-delta(Fig. 1) is close to, or even overlaps, the southern exten-sion of the Jordan Fault that is one of the active segmentsof the DST (Niemi and Ben-Avraham, 1994). The mouth ofthis fan-delta is located about 2 km from the revised locationof the epicenter of the 11 July 1927, Ms 5.9, Jericho earth-quake which supposedly occurred due to the reactivation ofthis fault (Shapira et al., 1993). The maximal distance fromthe Nahal Darga fan-delta to this epicenter will not exceed12 km even if it occurred at the farthest point from the fan-delta but within the error ellipse. This implies that a consider-able part of the deformation in this area should be associated


Jeru

sale

m -

Dea

d Se

a ro

ad

Nahal Darga

Dead Sea

N

1

2

3

4

2a

2b

0 250mSCALESeismic l ineGeol . cross sect ion

Fig. 2. The Nahal Darga fan-delta and its incised channel, the shaded area between (2a) and (2b). The outline of the seismic profiles isin heavy lines, and the numbers 1 to 4 represent the seismic profile lines 1 to 4. The southern geologic cross section of the channel walls,(a), and the northern one, (b), are outlined by thin line with arrows. The open circles on the channel walls represent the location of theexposed allochthonous bodies and the highly deformed blocks in the southern and northern channel walls respectively. The full circles onthe seismic profiles represent the first appearance, from west, of the large subsurface reverse faults and flower structures representing thewestern boundary of the eastern deformation zone (S.P. 100, 140 and 190 in profiles 2, and 3, and 4, respectively).

with seismic activity.

The Dead Sea lake level underwent significant changesduring the late Pleistocene and Holocene (e.g. Bowman,1971; Neev and Hall, 1977; Klein, 1982; and Frumkin et al.,1994). During the last 30 years the Dead Sea lake level waslowered at a mean rate of about 70 cm per year. This low-ering, by more than 20 m (Klein, 1982), is the consequenceof human intervention by using the sweet water of the DeadSea drainage system for human need. Due to this changein the Dead Sea lake level, the easternmost part of the fan-delta (Fig. 2) was incised by a 7–10 m deep channel exposingthe Late Pleistocene and Holocene section. A geological andstructural study, including detailed mapping of the channelwalls that was performed by Kadan (1997) is the basis of thegeological part of this study. It reveals the detailed structureof the fan-delta and the relationship between its gently eastdipping (up to 5◦–7◦) sedimentary units composed of con-glomerates, sandstones, silts and clays (Fig. 3, (Cadan et al.,1996; Enzel et al., 2000). The conglomerates represent de-position along the riverbed, or even within the lake in casesof floods. Very low dips of the conglomerate cross-sets char-acterize deposition along the riverbed, whereas high dips, upto a few tens of degrees (e.g. Fig. 3a, 370, unit 12), charac-

terize deposition in the lake depending on the water depth.The sandstones represent deposition along a beach environ-ment, and the silt and clays were deposited within the wa-ter lake. Due to the frequent changes of the Dead Sea lakelevel during the Pleistocene and Holocene, the deposition lo-cation of each type of sedimentary unit changed in space.During lake high stands sediments such as clays, carbonatesand silts, were deposited west of the present shoreline (e.g.Fig. 3a, unit 7, 115–160) because the lake extended west-ward. On the other hand during low stands, the depositionlocation of deltaic sandy gravely sediments was shifted tothe east (e.g. Fig. 3a, unit 12, 360–420) due to the eastwardretreat of the lake. In addition, field relationships indicatethat during lake level lowering major erosional unconformi-ties (e.g. Fig. 3b, 0–100; Fig. 3a, 140–200) were developed.At the eastern part of the fan-delta the regular, gently dippingcross-bedding sets change to very steep foresets of coarsegrained deposits (Gilbert Type sets (Fig. 3a, unit 12, 360–415), commonly characterizing deposition on steep slopes(Massari and Colella, 1988).


Fig. 3a. A continuous stratigraphic cross section along both channel walls of Nahal Darga fan-delta and Shot Point of the seismic profiles.Figure 3a presents the southern wall and 2b the northern one. The left end of each section of Fig. 3a is continued with the right end ofthe section below it, and the right end of each section of Fig. 3b is continued with the left end of the section below it. The lower scale isin meters; Numbers in circles represent stratigraphic units from the older to younger ones, and the heavy line is the major unconformitycrossing the study area. Important shot-points, drawn above or below the cross-section, represent the location of subsurface faults beneaththe stratigraphic cross section. The number of profile line appears in brackets and the shot-point beside it.The legend: 1. conglomerate; 2. Sandstone; 3. silt and shales; 4. cross-bedding; and 5. fault line. These stratigraphic cross sections are basedon tracing the main rock units on scaled photographs.

3 Faults and deformation zones in the Nahal Darga fan-delta

A full description of the various rock units comprising thestratigraphic section and the deformation features is foundelsewhere (Enzel et al., 2000). In this section we briefly de-scribe the structures that are relevant in examining the pos-sible association between the surface and subsurface struc-tures. The precise determination of surface faulting, evento a scale of a few centimeters, is facilitated by the vari-ous types of sediments with sharp bedding planes compris-ing the fan-delta and the intervening distinct unconformityplanes (Fig. 3). A detailed structural study reveals the exis-

tence of two deformation zones. The most dominant struc-tures in the western zone are two sets of faults (Fig. 3a, 75–115) whose mean strike is NNE-SSW and a mean dip 60◦–70◦ to WNW or ESE (Fig. 4). The displacement along thesefaults varies from a few centimeters to 2 m. A few faultscrosscut almost the entire section (e.g. Fig. 3b, 30), whereasthe majority are either covered by continuous unfaulted beds(e.g. Fig. 3a, 80–100), or affect only a few beds (e.g. Fig. 3b,275–285). Along some of the faults multiple faulting couldbe proved according to differential displacements along theirfault planes in which the displacement of the lower beds isconsistently higher than those of the upper strata (Enzel et


Fig. 3b. continued ...

al., 2000). Assuming that the faulting resulted from tectonicactivity, such multiple displacements indicate reactivation ofthe faults due to a few sequential earthquakes that occurred atdifferent times during the deposition of the fan-delta (Cadanet al., 1996). The faults in the eastern deformation zone areshort and cross only a few layers (e.g. Fig 3b, 255–285). Theeastern deformation zone is also characterized by a differentdeformation style. Three rock bodies, about 15 m wide, anda bowl-shape cross-section are exposed along the southernwall of the channel (Fig. 3a, 280–290, 300–307 and 327–331). A few layers including rotated and faulted blocks up toa few meters long and high (Fig. 3b, 287–320) are exposed inthe northern wall, NNE of the bowl-shape bodies only there.These bodies, on both sides of the channel, are restricted toone horizontal stratigraphic zone up to 5 m thick and are lo-cated above a continuous silty layer (unit 8, Fig. 3a) withno deformed beds above or below. Internal faults and de-formation features such as mixed layers (Marco and Agnon,1995) were observed in the rocks comprising the bowl-shapebodies but not in the adjacent rock of the fan-delta. Enzelet al. (2000) suggested that these allochthonous bodies are

most probably slumps and represent another style of defor-mation caused by medium to large earthquakes. However,the NNE-SSW trend, including the allochthonous bodies andthe rotated blocks, is parallel to the mean strike of all faultsin the fan-delta, and is located above sub-surface faults.

At least ten layers containing liquefaction structures wereobserved in the fan-delta within the fine-grained sandy andsilty lacustrine sediments. The scatter of the liquefied layersin the stratigraphic section is uneven, several groups of ad-jacent layers underwent liquefaction, whereas other groupswith similar composition but at different levels of the sec-tion, did not undergo such liquefaction. These liquefactionstructures indicate continuous seismic activity in this area.The groups of closely overlying liquefied layers indicate acluster of large earthquakes whereas the unliquefied groupsindicate quite period. Radiocarbon dating of part of the lique-fied strata suggests a mean recurrence frequency of approx-imately 600 years for magnitude>5.5 earthquakes (Cadanet al., 1996; Enzel et al., 2000). However, the faults withinthe two deformation zones could result from either displace-ments due to deep-seated faults or from gravitational slump-


N=33

N=36

Fig. 4b. Stereographic projection, lower hemisphere, of 99 poles(dots) to fault planes from Nahal Darga. The great circles are themean fault planes of the two groups of faults striking NNE-SSWwith dips to WNW and ESE. The small circles outline theα95range.

ing (Selby, 1982) of the sedimentary sequence as a responseto the existence of a deep basin to the east.

4 Shallow high resolution seismic study

4.1 Data acquisition

High-resolution seismic profiles were acquired in order tostudy the subsurface in the Nahal Darga fan-delta. Profiles1, 2, and 3 (Figs. 5a, 5b, 5c, respectively) were locatedalong the southern bank, the stream channel and the north-ern bank, respectively (Fig. 2). Profile 4 (Fig. 5d) is arcuatebecause it follows the Dead Sea coastline. The data acqui-sition system consisted of an EG&G recording system. Theseismic source was a Dynasource pneumatic hammer hittingthe ground, two shots per location, and the seismic line com-prised 10 Hz geophones and an off-end cable spread consist-ing of 48 channels with minimum offset of 5 m and 2.5 m be-tween receivers. The maximum common depth point (CDP)fold was 12 traces and source interval was 5 m. The totalrecording time was 1 s. with a sample rate of 0.0005 s.

4.2 Data processing

The processing sequence for all CDP stacked data was sim-ilar to conventional petroleum exploration processing flows,with the main distinctions related to scale and emphasis onevent identification after each processing step. Note that theterm “high resolution” is not only the recording parameters,but also the processing scheme of the recorded data; thusspecial attention during the processing step was paid to mute

determination. In particular, it is well known (Miller, 1992)that selection of improper normal moveout (NMO) mute onshallow reflection data can drastically degrade the quality ofCDP stacked sections. Stacking velocities were determinedthrough qualitative analysis of constant velocity stacks andwere followed by post-stack time migration.

Prestack depth migration would have been more appro-priate in this study but it requires an accurate and detailedknowledge of the shallow subsurface. Because of the lack ofthis information we did not use this process.

5 Results

We represent here only the results of the reflection wave pro-files, although exact determination of fault location withinthe profile is also based on diffraction waves. Usuallythe detection of faults on seismic sections is solved by vi-sual interpretation of stacked sections. However, when de-tecting faults with a vertical displacement of less than thewavelength, a more sophisticated method must be used. Inorder to obtain reliable information about possible struc-tural/lithological discontinuities from seismic data, a proce-dure utilizing certain features known to be associated withthe presence of such discontinuities is desirable. One suchspecific feature demonstrating the presence of faults on aseismic section is associated with the presence of diffractedwaves in the vicinity of the discontinuity location. Thesewaves can serve as a good indicator for fault detection. Weextracted diffractions from the unmigrated stacked sectionsand used this information for identifying real discontinuitiesin the subsurface. A detailed description of the method andexamples of its application may be found in Landa et al.(1987) and Bruner and Landa (1991).

The seismic interpretation is based on the presence of re-flection horizons which, most probably but not necessarily,imply stratigraphic correlation. The time is given in sec-onds by Two Way Time (TWT). Based on a priori informa-tion (Niemi and Ben-Avraham, 1997; Ben-Avraham, 1997;Kashai and Croker, 1987), average velocity in the investi-gated area is about 2000–2500 m/s, so the maximal informa-tion depth achieved by these seismic profiles is about 400 m.Profile 1(GI-047-EG; Fig. 5a): This profile is disturbed inits central part, (Shot Points (S.P.) 55–135) due to its proxim-ity to the steep southern wall of the stream channel (Fig. 2).Nevertheless, in its western part (S.P. 30–55) the subsurfacebedding dips gently eastward and one normal fault can bedetermined. This fault is located below the eastern part ofthe western deformation zone (Fig. 3a, 115). At the easternpart of this profile (S.P. 137–150) two normal faults are de-fined below the area in which the allochthonous bodies areexposed at the surface (Fig. 3a, 315).Profile 2(GI-048-EG; Fig. 5b): This profile was measured in the mid-dle of the Nahal Darga channel and, therefore, can be relatedto both channel walls. The western termination of this seis-mic line lies approximately at the eastern part of the west-


Fig. 5a. Uniterpreted and interpreted multichannel high-resolution seismic profiles from the Nahal Darga fan-delta. The location of theseprofiles is found in Fig. 2. The interpreted seismic markers do not have any stratigraphic correlation. The exact determination of fault locationwithin the profile is also based on diffraction waves. The time is given in seconds by Two Way Time (TWT). The vertical lines above theseismic profiles in each of the profiles relate to the crossing location of seismic profiles. For example, the vertical line at the eastern part ofprofile 1 is the location where profile 4 crosses profile 1. The three vertical lines above profile 4 are the locations where, from left to right,profiles 3, 2 and 1 cross profile 1 respectively.

ern deformation zone. No overlap exists between these twobecause an artificial waterfall, designed to protect the roadwhere it crosses the riverbed, prevented installation of thegeophysical equipment further to the west. Therefore, thefaults of the western deformation zone could not be detectedin this seismic profile. The subsurface bedding dips gentlytoward the east as expected of fan-delta layers which draineastwards. An overall subsurface positive flower structure,comprised of a major reverse fault associated with a cluster ofminor normal faults and a gentle fold, is outstanding betweenS.P. 100–130 and 0.05 s TWT. The location of these subsur-face faults is below the area of the allochthonous bodies inthe southern channel wall (Fig. 3a, 300) and the deformedblocks in the northern wall (Fig. 3b, 300). At S.P. 140 a nor-mal fault is observed in the subsurface extending almost tothe surface. This fault is located east of the eastern defor-

mation zone and underneath the surface area in which thebedding and cross-bedding become steep, Gilbert type sets(Fig. 3a, 395). Development of steeply dipping cross setsis commonly associated with deposition in deep water and,therefore, the existence of deep-seated normal faults beneaththese sets may suggest deepening of the lake near the coastdue to down faulting. At S.P. 65–80 and 0.06–0.1 s (TWT)a concave unconformity, probably the trace of an old smallchannel, is observed. This is similar to small erosional chan-nels within the surface section (e.g. Fig. 3a, 102, upper partof the section) or to the surface slumps.

Profile 3(GI-049-EG; Fig. 5c): This profile is characterized by in-tensive normal faulting, which at S.P. 45–80 forms a flowerstructure pattern with a gentle fold between the secondaryfaults at S.P. 80, 0.1 s TWT. The main subsurface fault, at


Fig. 5b. continued ...

S.P. 50, overlaps the intensively faulted area of the westerndeformation zone along the northern channel wall (Fig. 3b,30–50). Actually, this subsurface fault is found along thecontinuation of a surface fault that exhibits the largest ob-served vertical displacement in this fan-delta. At the easternpart of the profile a large reverse fault (its upper part at S.P.138, and lower at 157) which forms the western boundary ofa gentle anticline (its hinge at S.P. 160 and 0.07 s TWT) isobserved. This reverse fault is located just below the east-ern deformation zone in the northern bank of the channel(Fig. 3b, 275–314). Thickening and thinning of sedimentarylayers and a small erosional channel are seen at S.P. 140–200,0.05–0.15 s (TWT). A 25 m wide, bowl shaped unconformityis observed at S.P. 170 and 0.1 s (TWT) representing eitheran old buried channel or an allochthonous body.

Profile 4(GI-050-EG; Fig. 5d): This profile, measured along the DeadSea coast, includes three different domains. Flat bedding,and a few normal faults characterize the southern domain,up to S.P. 70. In the second domain (S.P. 70–190) the gen-eral subsurface pattern is that of a positive flower structurein which the strata are intensively deformed by many nor-mal and a few reverse faults and folds between the faults.The third domain (S.P. 195–305) is characterized by contin-uous bedding gently dipping southeastward. At S.P. 255 andtime 0.1 s (TWT) the strata are discontinuous and the bowlshape of the discontinuity is similar to the cross section ofthe exposed allochthonous bodies in the eastern deformationzone (e.g. Fig. 3a, 285 and 303 lower part), or an erosionalchannel. This similarity suggests a common mechanism ofdevelopment for the surface and subsurface stream erosionor allochthonous bodies. An unconformity including strati-graphic reduction of the thickness between two markers isobserved between S.P. 190–300 and time 0.1–0.12 s (TWT).There is a general consistency between the structures aroundthe intersection points of line 4 with lines 1, 2, and 3.


Fig. 5c. continued ...

Fig. 5d. continued ...


6 Discussion

Due to the constraints of shallow high-resolution seismicdata acquisition, reliable seismic data begin at a depth of ap-proximately 10 m and extend to about 400 m. This range ofseismic information results in a+10 m zone, immediatelybelow the surface, that lacks both surface geological informa-tion and seismic data. However, the proximity of the surfacedata to the subsurface data justifies the correlation betweenthe two. The interpreted stacked sections of profiles 1–3 (GI-047-EG, GI-048-EG, and GI-049-EG respectively) reveal ageneral eastward dip as expected of a fan-delta draining east-ward. In profile 4 (GI-050-EG), the dips are subhorizontal inthe southern part and have a SE component in the northernpart. These dips are expected from the NE trend of the south-ern part of this profile that is perpendicular to the trend of Na-hal Darga and the NW trend of the northern part of profile 4that is oblique to the Nahal Darga stream. Similar dip-trendsare seen, as expected, in the northern, eastern, and southernparts of the other fan-delta in this region. In addition to thesimilarity of surface and subsurface dips other geomorpho-logic features characteristic to fan-deltas, such as thinningand thickening of sediment layers and cut and feel smallchannels filled with younger sediments, are also observed.These observations indicate that the depositional conditionsthat prevailed in the fan-delta area during the Holocene alsoexisted in earlier times, while the subsurface sediments, tothe depth of 400 m, were deposited.

The main seismic reflectors determined in the profilesshould be bedding or unconformity planes as no other planeswere observed in the fan-delta exposed section. These reflec-tors are displaced along discrete planes, faults, similar to thedisplacements of the surface bedding, indicating that faultingis not limited to the upper part of the fan-delta but extendsdown to at least the Late-Pleistocene subsurface rock units.

It could be argued that displacements along the fault planesin the fan-delta section resulted from movements due to ten-sion cracks in the upper block of a landslide. Such landslidescould develop either as plane failure along dipping clay beds(e.g. Selby, 1982; Hoek and Bray, 1991) or as slumps alongshallow, spoon-shape, sliding planes (circular failure, Hoekand Bray, 1991), without any direct association to subsurfacefaults. The internal friction angle of soft sedimentary rockssuch as sandstone, chalk and shale is in the range of 25◦–35◦

(Hoek and Bray, 1991, Table 1). A smaller friction angle,16◦, was measured by Hatzor and Levin (1997) for clay filledbedding plane in limestones along which a plane failure oc-curred in a phosphate mine. However, the development ofthe faults within the Nahal Darga fan-delta due to landslidesis rejected because:

1. no shallow curved planes, along which slumping couldoccur, were detected in the seismic profiles to the depthof 400 m, and

2. the gentle dip of the bedding comprising the fan-delta,3◦–7◦, is too shallow to initiate sliding toward the lake.

A tectonic origin, most probably associated with seismicactivity along the DST faults, is suggested as the cause of thevarious structures developed in the fan-delta for the follow-ing reasons:

1. subsurface faults, displacing the reflectors in the buriedsection, extend from a depth of at least 400 m, which isthe depth of the seismic resolution, almost to the sur-face;

2. the good correlation observed in many cases betweensurface and subsurface faults suggests association be-tween them. A good example is the continuation be-tween the subsurface fault observed in the western partof profile 3 (with the reservation of the upper 10 m gapbetween the surface and subsurface data) with the sur-face fault along which the largest vertical displacementin this fan-delta was measured; and

3. the low inclination of available sliding planes (3◦–7◦)with respect to the available friction angle between thelayers (16◦–25◦).

Along several subsurface faults the displacement is limitedto the subsurface and does not continue upwards (e.g. pro-file 4, S.P. 220 and profile 3, S.P. 40). Such buried faults in-dicate old faults that were not reactivated by younger tectonicevents, similar to many surface faults, (e.g. Fig. 3a, 78–86).The thinning of layers toward the flower structure in profile5d, S.P. 300–190, is best explained by a small high at S.P.190, formed most probably due to faulting. This implies thatthe faulting and flower structure development has been ac-tive since the deposition of Nahal Darga section during thePleistocene.

The correlation between surface and subsurface faults isalso suggested by the different deformation style characteriz-ing the two surface deformation zones and their concurrencein the subsurface. The surface western deformation zone ofthe fan-delta is characterized by NNE trending normal faultswhereas that of the eastern zone is different and includes onlysmall faults confined to three layers (northern cross section,Fig. 3b, 268–285) and highly deformed strata (Fig. 3b, 280–320) and three NNE trending allochthonous bodies (slumps)(Fig. 3a, 280–350). The surface NNE trend of these al-lochthonous bodies is suggested by correlating them with thehighly deformed strata in the northern wall. The first appear-ance of the allochthonous bodies and the highly deformedstrata in the cross sections (Figs. 3a, 3b) is denoted by emptycircles on cross sections 3a and 3b (Fig. 3), respectively. ThisNNE trend is parallel with the mean strike of the small faultsin this area (Fig. 4). The deformation style of the fan-deltasubsurface also depends on the location along the fan-delta.The subsurface faults in the western part of the fan-delta aremainly normal (e.g. Fig. 5c, S.P. 50), whereas the seismic re-flectors in the eastern part include evidence of both normaland reverse faults in a positive flower structure pattern sug-gesting some compression along this zone.


The true strike of the subsurface faults is unknown, how-ever it could be estimated by comparing the trends of defor-mation zone boundaries at surface and subsurface. The firstappearance, from the west, of the large subsurface reversefaults and flower structures representing the western bound-ary of the eastern deformation zone (S.P. 100, 140 and 190 inprofiles 2, and 3, and 4 respectively) were marked on Fig. 2as full circles. The general trend of these circles, and hencethe trend of this boundary, is NNE and it is parallel with:

1. the mean strike of the exposed surface faults in the fan-delta (Fig. 4);

2. the line connecting the first appearance of the al-lochthonous bodies in the southern wall and the highlydeformed strata in the northern wall.

From Fig. 2 it is clear that the line of first appearance of theallochthonous bodies in the southern wall and the highly de-formed strata in the northern wall are very close to the line offirst appearance of the large reverse faults and flower struc-tures in the subsurface. This association strongly supportscorrelation between them. The existence of a positive flowerstructure fault pattern comprised of normal and reverse faultsand secondary folds (Twiss and Moores, 1992) in the subsur-face of the eastern deformation zone suggests tectonic activ-ity under a compressional component, although most surfacefaults are normal. These relationships may be explained byassuming N-S left-lateral movement along the NNE trend-ing eastern deformation zone. Such a left-lateral movementshould be attributed to tectonic activity along the Jordan fault(Fig. 1), or to one of its secondary splays that extends underthe eastern deformation zone of the fan-delta, or very closeto it. The compressional nature of the subsurface faulting inthe eastern part of the fan-delta also explains the differencebetween the nature of the western and eastern surface defor-mation zones observed in the fan-delta.

7 Conclusions

Using the Nahal Darga fan-delta case study we show howthe combination of a detailed geological mapping and a shal-low, high-resolution reflection study reveals the associationbetween surface deformation and subsurface faulting. Thisassociation is based on the co-existence of surface deforma-tion zones and subsurface major faults beneath them. Theserelationships exclude the possibility that the faults in the fan-delta resulted from incipient landsliding. The similarity ofsurface dips in different parts of the fan-delta and the geom-etry of the exposed unconformity planes and channels withtheir counterparts in the subsurface, suggest that Holocenesedimentation actually started in Late Pleistocene. The sur-face and subsurface deformation styles in the western defor-mation zone are different from that in the eastern one. Thesubsurface normal displacements below the western defor-mation zone and the positive flower structures in the sub-surface of the eastern one are in agreement with the sur-

face structures, normal faults in the west and tectonic al-lochthonous bodies (slumps) due to strike-slip movement inthe east. Thinning and thickening of sub-surface strata asso-ciated with fault and flower structures suggest that faultingwas already active during the Late Pleistocene. The presenceof a fault zone underneath the Gilbert Type (steep dipping,up to 20◦–30◦) sets suggests that the relatively deep water inwhich the steep sets were deposited probably resulted fromnear shore faulting.

It is suggested that high-resolution seismic data may beused in other tectonically active localities to distinguish be-tween faulting associated with gravitational mass movement(slumping, land slides and plane failures), and faulting dueto tectonics. Such an analysis and interpretation of cross sec-tions is essential in any attempt to calculate the recurrencetime of large earthquakes.

Acknowledgements.This study was supported by the Ministry forEnergy and Infrastructure. We are grateful to Prof. M. Gross,Dr. Y. Hatzor, Dr. Ch. Benjamini, Prof. A. Ginzburg, Ms. D. Artzi,and an anonymous reviewer for their critical reviews of this paper.

References

Amit, R., Zilberman, E., Porat, N., and Enzel, Y.: Geomorphicand paleoseismic evidence for large historical earthquake in theAvrona Playa – Dead Sea rift, Geol. Soc. Isr. Annual Meeting,Kefar Gila’di, 3, 1997.

Bartov, J.: A structural and paleogeographical study of the centralSinai faults and domes, Ph.D. Thesis, Hebrew Univ. Jerusalem,143p (in Hebrew, English abstr.), 1974.

Bell, W. B. and Katzer, T.: Surficial geology, hydrology, and lateQuaternary tectonics of the IXL Canyon area, Nevada. As relatedto the 1954 Dixie Valley earthquake, Nevada Bureau of Minesand Geology, Bull. 102, 52p, 1987.

Ben-Avraham, Z.: Structural framework of the Gulf of Elat(Aqaba), northern Red Sea, J. Geophys. Res., 90, 703–726, 1985.

Ben-Avraham, Z.: Geophysical framework of the Dead Sea: Struc-ture and Tectonics; The Dead Sea, Oxford Monographs on Geol.and Geophys., 36, 22–35, 1997.

Ben-Menahem, A.: Four thousands years of seismicity along theDead Sea Rift, J. Geophys. Res. 96(B12), 20 195–20 216, 1991.

Ben-Menahem, A., Nur, A., and Vered, M.: Tectonics, seismicityand structure of the Afro-Euroasian junction – The breaking ofan incoherent plate, Phys. Earth Planet. Inter., 12, 1–50, 1976.

Bowman, D.: Geomorphology of the shore terraces of the LatePleistocene Lisan Lake, Israel, Paleogeo. Paleoclim. Paleoecol.,9, 183–209, 1971.

Bowman, D.: Active surface ruptures on the northern Arava fault,the Dead Sea Rift, Isr. J. Earth Sci., 44, 51–52, 1995.

Bowman, D. and Gerson, R.: Morphology of the latest Quaternarysurface faulting in the Gulf of Elat region, eastern Sinai, Tectono-physics, 128, 97–119, 1986.

Bruner, I. and Landa, E.: Fault interpretation from high resolutionseismic data in the northern Negev, Israel, Geophysics, 56, 7,1064–1070, 1991.

Cadan, G., Bruner, I., Eyal, Y., Enzel, Y., and Landa, E.: Correlationbetween surface and subsurface structures at the Holocene fan-delta of Nahal Darga, Dead Sea, Geol. Soc. Isr. Annual Meeting,Elat, p. 16, 1996.


Enzel, Y., Amit, R., Bruce, J., Harrison, J., and Porat, N.: Mor-phologic dating of fault scarps and terrace risers in the southernArava, Israel: Comparison to other age-dating techniques andimplications for paleoseismicity, Isr. J. Earth Sci., 43, 91–103,1994.

Enzel, Y., Kadan, G., and Eyal, Y.: Holocene Earthquakes in theDead Sea Graben from a Fan-Delta Sequence, Quaternary Re-search, 53, 34–48, 2000.

Eyal, M., Eyal, Y., Bartov, Y., and Steinitz, G.: The tectonic devel-opment of the western margin of the Gulf of Elat (Aqaba) Rift,Tectonophysics, 80, 39–66, 1981.

Freund, R.: A model of the structural development of Israel andadjacent areas since Upper Cretaceous times, Geol. Mag., 102,189–205, 1965.

Freund, R., Garfunkel, Z., Zak, I., Goldberg, M., Weissbrod, T., andDerin, B.: The shear along the Dead Sea rift: Royal Society ofLondon Philosophical Transactions, ser. A, 267, 105–127, 1970.

Frumkin, A., Carmi, I., Zak, I., and Magaritz, M.: Middle Holoceneenvironmental change determined from the salt caves of MountSedom, Israel, in: Bar-Yosef, O. and Renee S. K. (Eds.):Late Quaternary Chronology and Paleoclimates of the EasternMediterranean, 315–332, 1994.

Gardosh, M., Reches, Z., and Garfunkel, Z.: Holocene tectonic de-formation along the western margins of the Dead Sea, Tectono-physics, 180, 123–137, 1990.

Garfunkel, Z.: Internal structure of the Dead Sea Leaky transform(rift) in relation to plate kinematics, Tectonophysics, 80, 81–108,1981.

Garfunkel, Z., Zak, I., and Freund, R.: Active faulting in the DeadSea Rift, Tectonophysics, 80, 1–26, 1981.

Gerson, R. and Grossman, S.: Later stages in the morphotectonicevolution of the southern Arava valley rift, Ministry of Energyand Infrastructure, Jerusalem. Rep. ES-1-91, 31pp, 1991.

Ginat, H., Eyal, Y., Bartov, Y., and Zilberman, E.: Mapping ofyoung, recent faults in the fluviatile fans of Elat, Geol. Surv. Isr.Rep., TR-GSI/14/94, 13pp, 1994.

Hatzor, Y. H.: Dynamic rock slope stability analysis at Masada na-tional monument using Block Theory and DDA, in: Amadei, B.,Kranz, R., Scott, G. A., and Smeallie, P. H. (Eds.): Rock Me-chanics for industry, Balkema, Rotterdam, 63–70, 1999.

Hatzor, Y. H. and Levin, M.: The shear strength of clay filled bed-ding planes in limestones – back analysis of a slope failure in aphosphate mine, Israel, Geological and Geotechnical Engineer-ing, 15, 263–282, 1997.

Heimann, A. and Ron, H.: Young faults in the Hula pull-apartBasin, central Dead Sea Transform, Tectonophysics, 141, 117–124, 1987.

Heimann, A., Zilberman, E., Amit, R., and Frieslander, U.: YesudHama’ala lineament – the main and young diagonal fault at thesouthern Hula Basin, Geol. Soc. Isr. Annual Meeting, KefarGil’adi, 45, 1997.

Hoek, E. and Bray, J. W.: Rock slope and Engineering, ElsevierApplied Science, London, 358pp, 1991.

Joffe, S. and Garfunkel, Z.: Plate kinematics of the circum Red Sea– a re-evaluation, Tectonophysics, 141, 5–22, 1987.

Kadan, G.: Evidence for Dead Sea Lake-Level Fluctuations andRecent Tectonism from the Holocene Fan-Delta of Nahal Darga,Israel, Unpublished M.Sc. thesis, Ben Gurion University of theNegev (in Hebrew with English Abstract), 1997.

Karcz, I., Kafri, U., and Meshel, Z.: Archaeological evidence forsub-recent seismic activity along the Dead Sea-Jordan rift, Na-ture, 269, 234–235, 1977.

Kashai, E. L. and Croker, P. F.: Structural Geometry and evolutionof the Dead Sea-Jordan rift system as deduced from new subsur-face data, Tectonophysics, 141, 33–60, 1987.

Klein, Z.: Morphological evidence of lake level changes, westernshore of the Dead Sea, Isr. J. Earth Sci., 31, 67–94, 1982.

Landa, E., Shtivelman, V., and Gelchinsky, B.: A method for de-tection of diffracted waves on common offset sections, Geophys.Prospecting, 35, 359–373, 1987.

Marco, S. and Agnon, A.: Prehistoric earthquake deformations nearMassada, Dead Sea graben, Geology, 23, 695–698, 1995.

Marco, S., Agnon, A., Ellenblum, R., Eidelman, A., Basson, U.,and Boas, A.: 817-year old walls offset sinistrally 2.1 m by theDead Sea Transform, Israel, Journal of Geodynamics, 24 (1–4),11–20, 1997.

Massari, F. and Colella, A.: Evolution and type of fan-delta systemsin some major tectonic setting, in: Nemec, W. and Steel, R. J.(Eds.): Fan-deltas: sedimentology and tectonic settings, Blackieand Son, London, 103–122, 1988.

Miller, R. D.: Normal moveout stretch mute on shallof reflectiondata, Geophysics, V. 57, 11, 1502–1507, 1992.

Neev, D. and Hall, J. K.: Climatic fluctuations during the Holoceneas reflected by the Dead Sea levels, International Conference onTerminal Lakes, Ogden, Utah, 5pp, 1977.

Niemi, T. M. and Ben-Avraham, Z.: Evidence for Jericho earth-quakes from slumped sediments of the Jordan River delta in theDead Sea, Geology, 22, 395–398, 1994.

Niemi, T. M. and Ben-Avraham, Z.: Active tectonics in the DeadSea basin, Oxford Monographs on Geol. And Geophys., 36, 73–81, 1997.

Reches, Z. and Hoexter, D. F.: Holocene seismic and tectonic activ-ity in the Dead Sea area, Tectonophysics, 80, 235–254, 1981.

Selby, M. J.: Hill slop materials and processes, Oxford UniversityPress, London, 264pp, 1982.

Shapira, A., Avni, R., and Nur, A.: A new estimate for the epicenterof the Jericho earthquake of July 11, 1927, Isr. J. of Earth Sci.,42, 93–96, 1993.

Sieh, K. E. and Jahns, R. H.: Holocene activity of the San Andreasfault at Wallace Creek, California, Geol. Soc. Am. Bull., 95, 8,883–896, 1984.

Twiss, R. J. and Moores, E. M.: Structural Geology, Freeman andCompany, New York, 1992.

Zak, I. and Freund, R.: Recent strike slip movements along theDead Sea Rift, Isr. J. Earth Sci., 15, 33–37, 1966.

Documents

High-resolution seismic study of the Nahal Darga fan-delta ...The left-lateral Dead Sea transform (DST) is an active plate boundary between the Arabian and the African plates (e.g